Comparison of metaheuristic optimization techniques for BWR fuel reloads pattern design

Fuel reload pattern optimization is a crucial fuel management activity in nuclear power reactors. Along the years, a lot of work has been done in this area. In particular, several metaheuristic optimization techniques have been applied with good results for boiling water reactors (BWRs). In this paper, a comparison of different metaheuristics: genetic algorithms, tabu search, recurrent neural networks and several ant colony optimization techniques, were applied, in order to evaluate their performance. The optimization of an equilibrium core of a BWR, loaded with mixed oxide fuel composed of plutonium and minor actinides, was selected to be optimized. Results show that the best average values are obtained with the recurrent neural networks technique, meanwhile the best fuel reload was obtained with tabu search. However, according to the number of objective functions evaluated, the two fastest optimization techniques are tabu search and Ant System.

► This paper shows a performance comparison of several optimization techniques for fuel reload in BWR.
► Genetic Algorithms, Neural Networks, Tabu Search and several Ant Algorithms were used.
► All optimization techniques were executed under same conditions: objective function and an equilibrium cycle.
► Fuel bundles with minor actinides were loaded into the core.
► Tabu search and Ant System were the best optimization technique for the studied problem.